03Housing in Britain and Europe from 1955 to 1979

Hypothesis 2: British housebuilding and housing outcomes of the post-war period between 1947 and 1980 were not typical of similar European countries

If Britain’s housing crisis only began after 1980, then we would expect to see its housing supply and outcomes in the post-war period until 1979 to be at least average for a European country.

Instead, evidence from the United Nations Housing and Construction Statistics actually shows that the UK built much less housing than almost all other Western European countries between 1955 and 1979. This deficit was the result of a uniquely low rate of private sector housebuilding which was not overcome by a relatively typical rate of public housing construction.

The gap remains even after accounting for population growth, demolitions, and the low quality of British housing stock, and meant British housing outcomes saw relative decline in the post-war period. The number of homes per person in Britain fell  from 5.5 per cent more homes per person than the average Western European country in 1955, to 1.8 per cent below the European average by 1979.

The UK has built less housing than other Western European countries since the mid-1950s

Figure 2 shows the average annual gross housebuilding rates from 1955 to 1979 for twelve Western European countries divided by tenure. The UK ranks towards the bottom of this list, seeing only 1.9 per cent growth in the number of homes every year, much less than France with 2.3 per cent annual growth in housing stock, and West Germany and the Netherlands on 3 per cent growth in housing stock every year. The data is described in Box 3.

Figure 2: Britain built much less than other European countries in the post-war era, especially private housing

Source: United Nations (all editions from 1957-2000), Annual Bulletin of Housing and Building Statistics for Europe (New York).

Differences in war damage seem not to explain this – Ireland, Switzerland, and Sweden built more than the UK despite being neutral in the Second World War.

Box 3: The United Nations’ Housing Data

The United Nations compiled yearly housing statistics from European governments from 1948 to 2000. The key variables for the analysis are:

Gross housebuilding:

The number of homes built every year is used to calculate the gross housebuilding rate. Where this is not available from the UN, values from Kohl (2017) have been taken instead.8 Incomplete data by country in the early years of this dataset means that 1955 is the earliest possible starting date for comparisons across most countries in the sample. When the data allows, 1948 is used as the starting date for as many individual comparisons with the UK as possible and noted as such.

Net increase in homes:

The number of homes within every country – and the net change in the stock over time – is provided by a different series within the UN dataset. Due to demolitions and conversions, the change in homes over time is not the same as the number of new homes that are built, which is instead captured by the housebuilding rate. This report distinguishes between the (net) number of homes added and the (gross) number of homes built to account for this.

Housing outcomes

The ratio of homes per person – the number of homes per thousand people – is used to compare housing outcomes between different countries in this paper. A higher number means the supply of housing is more abundant and people have more floorspace per person in their homes. The UK has a ‘head-start’ in housing outcomes over most European countries, in that it begins the post-war period in 1955 with more homes per person than the European average.

Comparing affordability internationally is not possible due to a lack of data within the UN dataset, and the difficulties of adjusting for different income levels, exchange rates, and variation within countries.

Tenure

This is divided into “public” or “private” housing based upon conventional definitions in the UK’s housing policy.

  • Public Housing: Housing built either by local authorities or non-profit corporations such as housing associations or the New Town development corporations. This includes housing built by private builders on a state contract to charge state-defined social rents and let it out to those eligible for social housing.
  • Private Housing: Housing built by for-profit builders or private individual self-build. This includes housing sold to new homeowners or privately rented for market rates that has had construction subsidised by the state.

After 2000, data for these variables has been collected from the relevant national statistical authorities.

Tenure mix:

The share of new supply that was delivered by either the private or public sectors. Some countries depended almost or primarily upon the private sector to build new housing in the postwar period, while other “mixed-tenure” countries had substantial roles for the public sector, including the UK.

Britain had the lowest private housebuilding rate in post-war Europe

Postwar Britain had a low rate of total housebuilding. Although the UK had a public sector housebuilding rate slightly above average for countries with significant public housing programmes,9 it had the lowest rate of private housebuilding in the post-war period of any Western European country.

A low rate of private housebuilding was not necessary to enable a large public housebuilding programme. To take two examples, from 1955 to 1979 both the Netherlands and Sweden had a higher total housebuilding rate than the UK. Their housebuilding rates by tenure and year that were in surplus of the UK can be seen in Figure 3. The Netherlands and Sweden built more private sector housing than the UK in almost every year from 1955-1979, and also had long periods in which they built more public sector housing than the UK.

Figure 3: The Netherlands and Sweden show that postwar Britain could have built more private and social housing

Source: United Nations, Bulletin of Housing and Construction.

The Netherlands and Sweden both indicate that the UK’s reliance on council housebuilding to deliver new homes in the post-war period was partly the result of low private housebuilding. Even though the Netherlands and Sweden had a smaller role for the public sector as a share of their tenure mix, both had higher average rates of public sector housebuilding at 1.4 per cent and 0.96 per cent respectively than the UK’s average rate of 0.9 per cent per year.

Neither population growth nor demolitions can explain Britain’s low housebuilding rates

Low gross housebuilding rates do not necessarily imply housing shortages. If Britain’s low population growth (including emigration) or a relatively low demolition rate explained low British gross housebuilding in the post-war period, then we would expect Britain’s net increase in dwellings to be much more typical of European countries after controlling for changes in the population.10

As Figure 4 shows, even after accounting for population growth and any impact from demolitions, the UK had one of the lowest increases in net housing supply in Western Europe from 1955-1979. For example, while the UK increased the number of homes available per person by 26 per cent, Switzerland managed an increase of nearly double this, at 48 per cent.

Figure 4: Post-war Britain’s increase in homes per person was low compared to other European countries

Source: United Nations, Bulletin of Housing and Construction. It is not possible to construct estimates for growth in the ratio of homes per person divided by tenure. *Belgian data is taken from 1955 until 1977.

West Germany stands out in Figure 4 for its exceptionally high growth in homes per person from 1955-1979, due to its need to both recover from extensive war damage and build extra housing for refugees. While West Germany’s number of net additions to the housing stock was unique, Switzerland and Sweden also both built much more than the UK after accounting for population growth despite being neutral in the Second World War.

Immigration to the UK cannot explain the low increase in per person housing stock in this period, as the UK experienced net emigration from the end of the Second World War until 1979, with average annual net migration running at -54,000 a year from 1964-1978.11 The effect of migration therefore increased the UK’s net growth in homes per person in Figure 4, and indicates that there were already housing supply issues in the UK before the shift towards net immigration in the 1980s and 1990s.

Postwar Britain lost its initial head-start in housing outcomes

An alternative explanation for low build rates during this period could be that Britain did not need to build as many new homes due to its head-start in housing outcomes over most European countries. British housing outcomes – the ratio of homes per person – were higher than the European average at the start of the post-war period in 1955.

Britain built less than other West European countries even after accounting for Britain’s housing head-start. Figure 5 shows the ratio of homes per person in different European countries in every year from 1955 to 1979 as a percentage of Britain’s ratio, which is fixed at 100 in every year. Countries that are above the line have more homes per person than the UK in that year, and below it they have fewer. Every country on the graph sees their housing outcomes improve relative to the UK.

Figure 5: The UK saw relative decline in housing outcomes over the post-war period

Source: United Nations, Bulletin of Housing and Construction.

West Germany’s rapid post-war recovery build rates can clearly be seen in Figure 4, with the country reaching British levels of homes per person by around 1967. West Germany’s housebuilding rate slows after this point, but the number of homes per person continues to increase beyond British levels. Similarly, Switzerland also manages to overtake the UK in homes per person around 1970.

Other countries such as Finland and the Netherlands started the period with far fewer homes per person than the UK, partly due to lower average incomes. As their economies and standard of living caught up with the UK from 1955 to 1979, they also saw homes per person rise towards British levels due to their extensive construction programmes, despite rapid population growth in the Dutch case.

Denmark and Sweden are different still – they began 1955 with more homes per person than the UK. Yet they managed to increase that lead further over the following two and a half decades, in Sweden’s case with a larger public housebuilding programme than the UK.

The result of British underbuilding during the post-war period was that the UK lost its initial head-start in housing outcomes. Although the UK’s number of homes per person was 5.5 per cent above the average Western European country in 1955, it had fallen to 1.8 per cent below the European average by 1979.

Postwar Britain’s low quantity of new homes were not high quality

Another argument might be that Postwar Britain’s poor housebuilding performance can be explained by a choice of “quality over quantity”, such as the influential Parker Morris space standards on council housebuilding.12 As the UK was building less than other European countries, if Postwar Britain had prioritised fewer but higher quality dwellings, then we should expect to see an average rate of investment in residential construction and larger dwellings than other European countries.

However, Figure 6 shows that the UK invested the least in housebuilding of any Western European nation as a share of GDP – only an average of 3.3 per cent of GDP per year, compared to 5.1 per cent in the Netherlands and 6.2 per cent in Finland.

Figure 6: Postwar Britain had the lowest investment rate in residential construction of any country in Europe

Source: United Nations, Bulletin of Housing and Construction. * Data for Switzerland only goes up to 1970 and data for Austria is from 1955 to 1960 and 1965 to 1966.

In addition, Table 1 shows that by 1992, average UK housing size by floor area was significantly lower than its peer countries in Western Europe. The average British dwelling– many of which will have been completed in the post-war period – was roughly equal in size to the average dwelling in Greece, which had much lower average incomes. The only average dwellings which were smaller than that of the UK were those in the former East Germany.

Table 1: The UK had the smallest dwellings in Europe soon after the end of the post-war period

Country Floorspace per dwelling, 1992 (m2)
Former East Germany 64.6
Greece* 79.6
United Kingdom 79.7
Austria 85
France 85.4
Former West Germany 86.6
Sweden 92
Ireland 95
Netherlands 98.6
Denmark 106.9

Source: Netherlands’ Ministry of Housing, Physical Planning and Environment (1992), “Statistics on Housing in the European Community.”

The quality of new stock also declined as the post-war era progressed, with a trend towards smaller properties and plots, using inferior materials and the removal of white goods from developer ‘bundles’ for new homeowners.13 Furthermore, new homes were in worse locations than existing stock, as the rapid expansion of green belts after 1955 meant that new houses on greenfield land were built further away from urban areas, with worse access to jobs and longer commutes than had pre-war trends continued.14

Britain’s housing outcomes have been falling behind Europe since the 1950s

The second hypothesis – that British housebuilding and housing outcomes of the post-war period between 1947 and 1980 were below similar European countries – is shown to be true. Britain built less housing and experienced relative decline in housing outcomes compared to peer European states over the post-war period before 1980.

Why the UK’s post-war housebuilding rate was so low requires an understanding of the policy choices the UK made and how they differed to the rest of Western Europe. These are the subject of the next section.

Footnotes

  • 8 https://www.sebastiankohl.com/data Dataset 4. “Residential construction, unbalanced panel, 1913-2016, 30 countries”.
  • 9 Excluding Switzerland and Belgium, which had a public housebuilding rate of around 0.1 per cent from 1947-1979.
  • 10 No good data on demolitions specifically exists by country, but the impact is captured within net changes to dwellings to person.
  • 11 ONS (2016), Long-Term International Migration into and out of the UK by citizenship, 1964 to 2015.
  • 12 Park, J. (2017), ‘One Hundred Years of Housing Space Standards’, Levitt Bernstein.
  • 13 Drewett, R. ‘The Developers: Decision Process’ in Hall ‘Containment of Urban England Vol 2.’, p.190. According to the 2001 English Housing condition survey semi-detached houses built between 1919-1944 had a floor area of 88.1 m2, while the values when built between 1945-64 and 1965-80 were 83.2 m2 and 83.3 m2 . For plot size the 1991 survey reports values of 355 m2, 341 m2 and 305 m2 for the respective periods. Holmans, A.E. (2005) ‘Historical Statistics of Housing in Britain’, Cambridge Housing and Planning Research, pp. 56-57.
  • 14 Hall P. and Thomas, R. and Drewett, R. “ “Urban Growth Towards a Verdict” in Hall ‘Containment of Urban England Vol 2’, pp. 381-382.